Water quality modeling for a tidal river network: A case study of the Suzhou River

Abstract

Combined with the basic characteristics of Suzhou plain river network, two modules are established, one of which is the hydrodynamic module using the water level node method involving gate operation, while the other is the water quality module based on the principle of WASP5 (water quality analysis simulation program5). These two modules were coupled and verified by the monitoring data of Suzhou River network. The results showed that calculation errors of {\mathrm{NH}}_4^{+}-\mathrm{N} and DO for the model were in the ranges of –15%—13% and –18%—16%, respectively. Despite of the deviations between the monitoring data and simulation result, the calculation accuracy of the model conforms to the practical engineering requirement. Therefore, the proposed coupling model may be useful for water quality simulation and assessment for river network under tidal influences.

Keywords

water quality model / coupling model / river network

Introduction

The Suzhou River, which runs through the Shanghai area, is a major tributary of the Huangpu River. It is a tidal river belonging to the lake source type, whose bound can be traced back to the upper reach of Huangdu. As one of the main surface water sources for Shanghai City, it has been playing an important role in the drainage, navigation and irrigation (Chang et al., 2010; Hua et al., 2006). Since the mid-20th century, the Suzhou River has been contaminated by a large amount of industrial and domestic sewage from the coast with the rapid economic development of Shanghai, leading to serious water environmental degradation and pollutant accumulation in the sediment.

Since the 1980s, rehabilitation work for the Suzhou River has gradually progressed. After the first two stages of reconstruction, the water quality of Suzhou plain river network has significantly improved. This paper is intended to simulate the environment of Suzhou River by using WASP5 (water quality analysis simulation program5) to establish the water quality model. With pollutant inflows and outflows controlled, hopefully, proper advice can be provided for the decision-makers (Liu et al., 2009; Zhu et al., 2011).

Methodology

Hydrodynamic model

With reference to the specific characteristics of water environment in the Suzhou plain river network, we establish a one-dimensional hydrodynamic model with Saint-Venant equations as the controlling and dispersing equations by using Preissmann four-point implicit difference scheme, and use the river-node method to solve the discrete equations. Then the necessary hydrodynamic data that we need can be obtained from the established model (Lai and Wang, 2002).

The Saint-Venant governing equations are employed to describe the one-dimensional unsteady flow:

(1){\displaystyle \frac{\partial Z}{\partial t}}+{\displaystyle \frac{\mathrm{1}}{B}}{\displaystyle \frac{\partial Q}{\partial x}}=\mathrm{0},

(2){\displaystyle \frac{\partial Q}{\partial t}}+{\displaystyle \frac{\partial }{\partial x}}\left(\alpha {\displaystyle \frac{Q^{\mathrm{2}}}{A}}\right)+gA\left({\displaystyle \frac{\partial Z}{\partial x}}+{\displaystyle \frac{Q\left|Q\right|}{K^{\mathrm{2}}}}\right)=\mathrm{0},

where Z represents the water level (m); B is the river surface width (m); Q is the flow (m³/s); x stands for space coordinates; t is the time coordinates; α is the momentum correction factor; g is the gravity acceleration; A is the cross section water area (m²); and K is the unit discharge.

Dispersed by the Preissmann four-point implicit finite difference, we obtain the one-dimensional nonlinear equations from the Saint-Venant equation. Then the linear algebra equations are solved with the river-node method. The internal boundary condition refers to the hydraulic structures including the sluice gate, bridge, pump station and lateral inflow. According to the features of the structures, the boundary condition should be given according to the features of the structures in line with the compatibility condition when the construct of hydraulic characteristics is not compatible with the Saint-Venant equation (Ambrose et al., 1993b; Paredes et al., 2010).

Water quality model

WASP5 is a dynamic compartment model that can be used to analyze a variety of water quality problems in such diverse water bodies as ponds, streams, lakes, reservoirs, rivers, estuaries, and coastal waters. The equations solved by WASP5 are based on the key principle of the mass conservation, which requires that the mass of each water quality constituent being investigated must be accounted for in one way or another. WASP5 traces each water quality constituent from the point of spatial and temporal input to its final point of export, conserving mass in space and time. By expanding the infinitesimally small control volumes into larger adjoining “segments” and by specifying proper transport, transformation parameters and loading, WASP5 implements difference equations of three-dimensional mass transportation. For brevity and clarity, however, considering the concerned river as a one-dimensional reach with an assumption of vertical and lateral homogeneity, the partial derivative of mass balance equation is presented as (Ambrose et al., 1993a; Fitzpatrick, 2009):

(3){\displaystyle \frac{\partial }{\partial t}}(AC)={\displaystyle \frac{\partial }{\partial x}}(-U_{x}AC+E_{x}A{\displaystyle \frac{\partial C}{\partial x}})+A(S_L+S_B)+A{S}_K,

where C is the concentration of the water quality constituent (mg/L or g/m³); t is the time (day); U_x is the longitudinal velocities (m/d); E_x is the longitudinal diffusion coefficients (m²/d); S_L is the direct and diffuse loading rate (g/(m³·d)); S_B is the boundary loading rate, including upstream, downstream, benthic, and atmospheric (g/(m³·d)); S_k is the total kinetic transformation rate (g/(m³·d)), positive one means source while negative one implies sink; A is the cross-sectional area (m²). The proposed model takes water quality indices including {\mathrm{NH}}_4^{+}-N , NO₃-N, PO₄, PHYT, CBOD, DO, ON, and OP into account. Also, the time series of water quality indicators, C_i(t), can be regarded as the boundary condition of the single inflow river section.

Case study

Background

There are 120 rivers, 71 river nodes, 20 boundary nodes (3 for water level and 17 for water discharge) and 662 cross sections in the Suzhou River network. Huangdu water level in the upstream Suzhou River, Huangpu Park and Wusongkou water levels are chosen as the boundary conditions of river network. The time series C_i(t) as the water quality indices are considered as the boundary condition of the water quality simulation. The calculated river layout is shown in Fig. 1.

Hydrological data

The model is tested and verified by the synchronization monitoring data, hydrological data from 23:00, April 1st to 23:00, April 11st, 2000. The roughness of the river network is set between 0.020 to 0.030 by calculating and debugging, and the monitored pollution source of boundary water level is Huangdu, Huangpu Park and Wusongkou. For it was rainy from April 1st to 16th, the measured data from the rainfall stations in Qingpu and Songjiang were used in the runoff calculations. During the water transfer, we incorporate the running of 25 sluice gates including Wenxi Gate, Yantietang Gate, Fengbang Gate, Xinchapu Gate, Huacao Gate and Pengyuepu Gate, etc.

Water quality monitoring data

There are 17 water quality monitoring sections during the water transfer process. The major monitoring indices include water temperature, COD_Cr, BOD₅, {\mathrm{NH}}_4^{+}-N, DO, water chroma and so on. The location of the monitoring sections is shown in Fig. 1.

Pollution source data

Testified data of the model were obtained from the pollution resource field investigation in 2000. According to the survey report, industrial emissions, enterprises discharge, living water discharge, livestock emission and the catchment of the surface runoff pollution are included in the Suzhou River network in Shanghai (Li et al., 2005). Based on the resources census data in Shanghai, we condensed the pollution sources into 17-point sources within the catchment, which were discharged into the corresponding rivers.

Water quality parameters

The model verified the water quality indices covering {\mathrm{NH}}_4^{+}-N, COD_Cr, BOD₅ and DO. The calculation coefficients of water quality simulation are: diffusion coefficient E_x value varies from 150 to 450 m²/d, aerobic coefficient in 20°C k_d varies from 0.15 to 0.35 d^-1, temperature coefficient θ_d is 1.05, half-saturation coefficient of COD k_COD is 0.4 mg/L, 20°C Nitrification coefficient k₁₂ is 0.09-0.13d^-1, temperature coefficient θ₁₂ is 1.08, ammonia half-saturation coefficient k_NIT is 0.5 mg/L, reaeration coefficient k₂ varies from 4.1 to 4.7 d^-1, and the temperature coefficient θ₂ is 1.02826 (Villegas and Giner, 1973; Ruan, 2000).

Summary

The results of the hydrodynamic model are compared with the monitoring data at Pengyuepu Gate, Beixinjing Gate, and Zhejiang Road Bridge. As shown in Fig. 2, the simulated discharge and monitoring one fits quite well, with the average error less than 24%, which may result from the potential measuring error and oversimplified section.

The comparisons of {\mathrm{NH}}_4^{+}-N and DO at Pengyuepu sluice gate, Beixinjing sluice gate and Zhejiang Road Bridge are shown in Figs. 3 and 4. The results showed that calculation errors of {\mathrm{NH}}_4^{+}-N and DO of the model were in the ranges of -15%—13% and -18%—16%, respectively, implying that the accuracy of the model conformed to the practical engineering requirement.

The proposed model can simulate the change of water flow and quality in the river network with hydraulic engineering and tidal influences. Since the finite difference discrete equation of water quality has the advantage of stable convergence, the application of such model may provide scientific basis for the function division of river network, sewage emission control and planning of water environment capacity. It may also be used as a powerful tool for the integrated river basin management if the socio-economic factors are incorporated into the model.

References

[1] Ambrose R B, Wool T A, Martin J L (1993a). The Water Quality Simulation Program, WASP5: Part A:Model Documentation. Environmental Research Laboratory, Athens, Georgia 30613

[2] Ambrose R B, Wool T A, Martin J L (1993b). The Water Quality Analysis Simulation Program, WASP5, Part B: The WASP5 Input Dataset. Environmental Research Laboratory, Athens, Georgia 30605

[3] Chang W T, Wang G, Han L X (2010). Inverse study on pollution source in tidal river on the basis of 2D water quality model. Water Resource Protection, 6: 5–8

[4] Fitzpatrick J J (2009). Assessing skill of estuarine and coastal eutrophication models for water quality managers. Journal of Marine System, 76 (1-2): 195–211

[5] Hua Z L, Han L X, Yao Q (2006). Environmental Hydraulics and Its Application. Nanjing: Hohai University Press

[6] Lai X J, Wang D G (2002). 1-D and 2-D coupling numerical model of unsteady flow. Science and Engineering, 6: 48–51

[7] Li G Y, Lu J, He Y J (2005). Relation between diversity of phytoplankton and environmental factors in the Jiaozhou Bay. Trends of Marine Geology, 4: 10–13

[8] Liu D Y, Lin W P, Zhao M (2009). Community structural characteristics of phytoplankton in Suzhou creek. Resource and Environment in the Yangtze Basin, 10 (18): 914–918

[9] Paredes J, Andreu J, Solera A (2010). A decision support system for water quality issues in the Manzanares River (Madrid, Spain). Science of the Total Environment, 408: 2576–2589

[10] Ruan Y L (2000). Shanghai Water Environment Research. Beijing: Science Press, 65–67

[11] Villegas I, de Giner J (1973). Phytoplankton as a biological indicator of water quality. Water Resource, 7(3): 479–487

[12] Zhu X F, Wang J D, Solo-Gabriele H M, Fleming L E (2011). A water quality modeling study of non-point sources at recreational marine beaches. Water Research, 45: 2985–2995

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg